Application of gain-scheduled vibration control to nonlinear journal-bearing supported rotor

Author(s)

Alves, Diogo Stuani ^[1] ; Wu, Matheus Freire ^[1] ; Cavalca, Katia Lucchesi ^[1]

Affiliation

^[1] Univ Estadual Campinas, Lab Rotating Machinery, Sch Mechan Engn, 200 Rua Mendeleyev, BR-13083860 Campinas, SP - Brazil

• Document type: Journal article
• Source: Journal of Sound and Vibration; v. 442, p. 714-737, MAR 3 2019.
• Web of Science Citations: 1

Abstract

Bearing reaction forces are commonly approximated by linear coefficients at a given rotational speed, although they are, in reality, considerably nonlinear in some situations. Therefore, a precise numerical simulation is highly desirable before experimental tests, usually resorting to integration of classical Reynolds equation, a well-known numerically time consuming. Due to faults or unusual demands, controllers can be applied to avoid failure. However, control action changes the machine dynamic behavior. This paper investigates the performance of a faster simulation method using a high order Taylor force approximation, by means of nonlinear coefficients, in situations that differs from the one in which the force was originally identified. This is accomplished using the force generated by a gain-scheduled static H-infinity controller. The proposed controller aims to stabilize fluid-induced instability and attenuate unbalance vibration at critical speed of a journal bearing supported rotor. A polynomial fit is applied to express the hydrodynamic bearings coefficients according to rotor rotational speed. The resultant polynomial system is then used in linear matrix inequalities from a two-stage method in order to synthetize a controller linearly dependent on rotational speed. Furthermore, the control efficiency is verified under the influence of bearing nonlinearities. The numerical simulations were compared with experimental data measured in a laboratory test rig. The results reveal that the control is effective in reduce vibration amplitudes especially in critical speed and fluid induced instability, and that the nonlinear coefficients are promising for approximating journal motion for both uncontrolled and controlled systems. (C) 2018 Elsevier Ltd. All rights reserved. (AU)

• FAPESP's process: 16/13059-1 - Linear parameter varying control applied to rotating machinery
• Grantee: Matheus Freire Wu
• Support Opportunities: Scholarships in Brazil - Doctorate (Direct)
• FAPESP's process: 15/20363-6 - Fault tolerant identification and control of rotating systems
• Grantee: Katia Lucchesi Cavalca Dedini
• Support Opportunities: Research Projects - Thematic Grants